Sanitation operation management and control method, system and device and medium

Abstract

The environmental sanitation operation management and control method comprises the following steps: constructing a target knowledge graph based on historical data and / or environmental data; processing the target knowledge graph based on a large language model, and determining a job scheduling sequence; and processing the job scheduling sequence based on an optimization algorithm, and determining an optimal job scheme. The environmental sanitation operation data can be dynamically adjusted based on the data updated in real time, and the operation effect can be dynamically monitored.

Description

Technical Field

[0001] This specification relates to the field of sanitation scheduling and control technology, and in particular to a sanitation operation control method, system, device and medium. Background Art

[0002] With the advancement of smart city construction, the demand for intelligent sanitation operations is growing. Traditional sanitation dispatch systems rely primarily on fixed operation templates and manual inspections, making it difficult to cope with dynamically changing environmental factors and complex anomaly detection requirements.

[0003] Therefore, it is necessary to provide an improved sanitation operation management and control method, system, device and medium, which can dynamically adjust sanitation operation data based on real-time updated data and dynamically monitor operation results. Summary of the Invention

[0004] One or more embodiments of the present specification provide a method for sanitation operation management and control, including: constructing a target knowledge graph based on historical data and / or environmental data; determining a job scheduling sequence by processing the target knowledge graph based on a large language model; and determining an optimal job plan by processing the job scheduling sequence based on an optimization algorithm.

[0005] In some embodiments, the target knowledge graph includes one or more of a static graph, a dynamic graph, and an experience graph.

[0006] In some embodiments, the processing of the job scheduling sequence based on the optimization algorithm to determine the optimal job plan includes: determining the event urgency, resource consumption and abnormality probability corresponding to the job scheduling sequence; based on the event urgency, resource consumption and abnormality probability, determining the reward value corresponding to the job scheduling sequence through the optimization algorithm; and determining the optimal job plan based on the reward value.

[0007] In some embodiments, the method further includes: obtaining vibration data of the operating vehicle; determining the probability of false operation based on processing of the vibration data by the LSTM network; obtaining images before and after the operation, and determining the similarity of the images before and after the operation; and determining whether to trigger an early warning based on the similarity of the images before and after the operation and the probability of false operation.

[0008] In some embodiments, the method further includes: obtaining the operation time of the operation vehicle; and determining whether to trigger an early warning based on the operation time, the similarity between the images before and after the operation, and the false operation probability.

[0009] In some embodiments, the method further includes: updating the target knowledge graph based on real-time data; determining whether to adjust the operation rules and whether to adjust the optimal operation plan based on the updated target knowledge graph and preset trigger rules.

[0010] In some embodiments, the method further includes: in response to the need to adjust the optimal job plan: determining the updated job scheduling sequence based on the processing of the updated target knowledge graph by the large language model; and determining the updated optimal job plan based on the processing of the updated job scheduling sequence by the optimization algorithm.

[0011] At the same time, one or more embodiments of the specification provide a sanitation operation management and control system, including a graph construction module, configured to: construct a target knowledge graph based on historical data and / or environmental data; a model processing module, configured to: determine a job scheduling sequence based on processing of the target knowledge graph by a large language model; and a solution determination module, configured to: determine an optimal job solution based on processing of the job scheduling sequence by an optimization algorithm.

[0012] One or more embodiments of the present specification provide a sanitation operation control device, including a processor, wherein the processor is used to execute the sanitation operation control method.

[0013] One or more embodiments of the present specification provide a computer-readable storage medium, wherein the storage medium stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes the sanitation operation management and control method. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] This specification will be further described in the form of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting, and in these embodiments, like numbers represent like structures, wherein:

[0015] Figure 1 is a schematic diagram of a sanitation operation management and control system according to some embodiments of this specification;

[0016] Figure 2 is an exemplary flow chart of a sanitation operation management and control method according to some embodiments of this specification;

[0017] Figure 3 This is an exemplary schematic diagram of determining whether to trigger an early warning according to some embodiments of this specification. DETAILED DESCRIPTION

[0018] To more clearly illustrate the technical solutions of the embodiments of this specification, the following briefly describes the drawings required for describing the embodiments. Obviously, the drawings described below are merely examples or embodiments of this specification. Those skilled in the art can apply this specification to other similar scenarios based on these drawings without inventive effort. Unless otherwise apparent from the context or otherwise noted, the same reference numerals in the figures represent the same structure or operation.

[0019] It should be understood that the terms "system," "device," "unit," and / or "module" used herein are a method for distinguishing different components, elements, parts, portions, or assemblies at different levels. However, if other terms can achieve the same purpose, the terms may be replaced by other expressions.

[0020] As used in this specification and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not refer to the singular but also include the plural. Generally speaking, the terms "comprises" and "include" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.

[0021] Flowcharts are used throughout this specification to illustrate the operations performed by systems according to embodiments of this specification. It should be understood that preceding or following operations do not necessarily need to be performed in exact order. Instead, the steps may be processed in reverse order or simultaneously. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0022] Figure 1 It is an exemplary module diagram of the sanitation operation management and control system shown in some embodiments of this specification.

[0023] In some embodiments, as Figure 1 As shown, the sanitation operation management and control system 100 may include a graph construction module 110 , a model processing module 120 , and a solution determination module 130 .

[0024] In some embodiments, the graph construction module 110 is configured to construct a target knowledge graph based on historical data and / or environmental data.

[0025] In some embodiments, the target knowledge graph includes one or more of a static graph, a dynamic graph, and an experience graph. For more information about graphs, see Figure 2 The corresponding content.

[0026] In some embodiments, the model processing module 120 is configured to determine a job scheduling sequence based on processing of the target knowledge graph by a large language model.

[0027] In some embodiments, the solution determination module 130 is configured to determine an optimal job solution based on processing the job scheduling sequence by an optimization algorithm.

[0028] In some embodiments, the solution determination module 130 is further configured to determine the event urgency, resource consumption and abnormal probability corresponding to the job scheduling sequence; based on the event urgency, resource consumption and abnormal probability, determine the reward value corresponding to the job scheduling sequence through the optimization algorithm; and determine the optimal job solution based on the reward value.

[0029] In some embodiments, the sanitation operation management and control system 100 also includes a judgment module (not shown in the figure), which is configured to obtain vibration data of the operation vehicle; determine the probability of false operation based on the processing of the vibration data by the LSTM network; obtain images before and after the operation, and determine the similarity of the images before and after the operation; based on the similarity of the images before and after the operation and the probability of false operation, determine whether to trigger an early warning.

[0030] In some embodiments, the judgment module is further configured to obtain the operation time of the operation vehicle; and determine whether to trigger an early warning based on the operation time, the similarity of the images before and after the operation, and the false operation probability.

[0031] In some embodiments, the judgment module is further configured to update the target knowledge graph based on real-time data; based on the updated target knowledge graph and preset trigger rules, determine whether to adjust the operation rules and whether to adjust the optimal operation plan.

[0032] In some embodiments, the sanitation operation management and control system 100 further includes an early warning module (not shown in the figure), which is configured to issue an early warning in a preset form in response to a need to trigger an early warning.

[0033] In some embodiments, the sanitation operation management and control system 100 also includes an update module (not shown in the figure), which is configured to adjust the optimal operation plan in response to the need: based on the large language model processing of the updated target knowledge graph, determine the updated operation scheduling sequence; based on the optimization algorithm processing of the updated operation scheduling sequence, determine the updated optimal operation plan.

[0034] In some embodiments, the various modules of the sanitation operation management and control system can be fully or partially integrated into a processor to implement the corresponding functions. In some embodiments, the processor can be a single server or a server group. The server group can be centralized or distributed. In some embodiments, the processor can be local or remote. In some embodiments, the processor can be implemented on a cloud platform.

[0035] It should be noted that the above description of the sanitation operation control system and its modules is for convenience only and does not limit this specification to the scope of the embodiments. It is understandable that those skilled in the art, after understanding the principles of the system, may arbitrarily combine the modules or form subsystems connected with other modules without deviating from the principles. In some embodiments, Figure 1 The atlas construction module 110, model processing module 120, and solution determination module 130 can be different modules within a single system, or a single module can implement the functions of two or more of the aforementioned modules. For example, the modules can share a storage module, or each module can have its own storage module. Such variations are within the scope of protection of this specification.

[0036] Figure 2 This is an exemplary flow chart of the sanitation operation control method according to some embodiments of this specification. Figure 2 As shown, the process 200 includes the following steps. In some embodiments, the process 200 can be executed by the sanitation operation control system 100 or a processor.

[0037] Step 210: Build a target knowledge graph based on historical data and / or environmental data.

[0038] The target knowledge graph refers to a structured semantic network that is used to represent entities (e.g., roads, weather events) and their relations (e.g., “roads are affected by weather”).

[0039] In some embodiments, the target knowledge graph includes one or more of a static graph, a dynamic graph, and an experience graph.

[0040] Static maps are used to store fixed data, which may include data that does not change easily, such as road grades and trash bin distribution. In some embodiments, static maps may be updated periodically based on a certain period.

[0041] As an example only, a static map can store basic municipal data, including road grades, school / commercial area coordinates, road materials, garbage bin distribution density, statutory operation time constraints, etc.

[0042] Dynamic graphs are used to store data that needs to be updated in real time, such as severe weather warning data, traffic congestion information, etc. Dynamic graphs can be updated in real time based on the real-time updated data.

[0043] As an example only, dynamic graphs can access external API data in real time to store data such as precipitation warnings from the Meteorological Bureau, congestion indexes from traffic departments, and emergencies (road construction).

[0044] The experience graph is used to store rule data extracted from historical work orders, such as delayed collection and delivery in commercial areas on Fridays. The experience graph can be updated regularly based on updates to historical work orders.

[0045] In some embodiments, the experience graph can be generated based on TF-IDF (Term Frequency-Inverse Document Frequency), a text mining technique used to assess the importance of words in a document. Based on TF-IDF, key rules can be extracted from historical work order texts to generate the experience graph.

[0046] As an example only, the experience graph may include implicit rules mined from historical work order texts (such as "delayed collection and transportation on the commercial street on Friday nights"), and by using the TF-IDF algorithm to extract key experiences, quantifiable scheduling constraints are formed.

[0047] Step 220: Determine the job scheduling sequence based on the processing of the target knowledge graph by the large language model.

[0048] In some embodiments, the large language model can be a fine-tuned LLM model, etc. In some embodiments, by inputting the target knowledge graph into the large language model, the large language model processes the fixed data, real-time data, and experience data therein, and based on the required instructions for outputting the job scheduling sequence, a corresponding job scheduling sequence can be output.

[0049] In some embodiments, the large language model may be required to output multiple job scheduling sequences based on different emphases.

[0050] Step 230: Determine the optimal operation plan based on the processing of the operation scheduling sequence by the optimization algorithm.

[0051] In some embodiments, the processing of the job scheduling sequence based on the optimization algorithm to determine the optimal job plan includes: determining the event urgency, resource consumption and abnormal probability corresponding to the job scheduling sequence; based on the event urgency, resource consumption and abnormal probability, determining the reward value corresponding to the job scheduling sequence through the optimization algorithm; and determining the optimal job plan based on the reward value.

[0052] In some embodiments, the optimization algorithm may be a Markov Decision Process (MDP), which is an algorithm used to model sequential decision problems in a dynamic environment.

[0053] In some embodiments, the processor can determine the event urgency, resource consumption, and abnormality probability corresponding to each job scheduling sequence based on historical data. The abnormality probability refers to the probability that the job result does not meet the job standard when the job scheduling is executed. It can be determined based on the total number of executions of the job scheduling sequence in historical jobs and the number of qualified times. In some embodiments, the optimization algorithm can include a reward function. Exemplarily, the reward function is as follows:

[0054] Q = a*xb*yc*z, where Q is the reward value corresponding to each job scheduling sequence, x, y, and z are the event urgency, resource consumption, and exception probability corresponding to each job scheduling sequence, respectively, and a, b, and c are the weights corresponding to the event urgency, resource consumption, and exception probability corresponding to each job scheduling sequence, respectively. These can be adjusted or preset based on historical experience. In some embodiments, a = 0.6 ± 0.1, b = 0.3 ± 0.05, and c = 0.1 ± 0.02.

[0055] In some embodiments, the processor may generate an optimal work plan based on the work scheduling sequence with the highest reward value and send it to the corresponding work vehicle.

[0056] In the sanitation operation management and control method of the present invention, the dynamic adaptability of the formulated operation plan can be improved by updating the target knowledge graph. By balancing the weights of event urgency, resource consumption and abnormal probability, resource optimization can be achieved and operation efficiency and effectiveness can be improved.

[0057] Figure 3 This is an exemplary schematic diagram of determining whether to trigger an early warning according to some embodiments of this specification.

[0058] like Figure 3 As shown, the processor can obtain vibration data 310 of the working vehicle; determine the probability of false operation 320 based on the processing of the vibration data by the LSTM network; obtain images before and after the operation 330, and determine the similarity of the images before and after the operation 340; based on the similarity of the images before and after the operation and the probability of false operation, determine whether to trigger an early warning.

[0059] In some embodiments, the processor can identify abnormal operating modes, such as "broom idling," based on vibration data acquired by a triaxial accelerometer mounted on the work vehicle. A triaxial accelerometer is a hardware device that can measure acceleration in three-dimensional space and can be used to acquire corresponding sensor data.

[0060] In some embodiments, if the estimated false operation probability is higher (eg, the number of occurrences of the abnormal operation mode exceeds a first threshold), the processor determines that an early warning needs to be triggered.

[0061] In some embodiments, the processor can determine the similarity between the images before and after the operation based on the SSIM index corresponding to the images before and after the operation, thereby obtaining the change in road cleanliness before and after the operation. The SSIM index is a metric that measures the similarity between two images and comprehensively considers brightness, contrast, and structural information.

[0062] In some embodiments, the processor may determine the change in road cleanliness before and after the operation based on ΔS=1−SSIM, where SSIM is the similarity of the images before and after the operation.

[0063] In some embodiments, if the change in road cleanliness before and after the operation is lower (eg, lower than a second threshold), the processor deems it necessary to trigger an early warning.

[0064] In some embodiments, the processor may further obtain the operation time of the operation vehicle; and determine whether to trigger an early warning based on the operation time, the similarity between the images before and after the operation, and the false operation probability.

[0065] As an example only, if a false operation occurs, and if the roadside camera captures images before and after the operation, the structural similarity (SSIM) is calculated based on the CLIP model. If the cleanliness change ΔS is less than 15%, and at the same time, the operation time is less than 20% of the historical shortest time for the same road section, the processor considers that an alarm needs to be triggered.

[0066] Among them, if vibration data is collected based on the vehicle-mounted three-axis acceleration sensor, the frequency domain energy is analyzed through the LSTM network, and the energy corresponding to the vibration data is identified to be less than 70% of the historical median, it is considered that false operations such as "broom idling" have occurred.

[0067] The present invention uses multimodal data cross-validation, combines equipment sensors with visual data, improves the recognition rate of false operations, solves the problem of single signal fraud, and enhances the accuracy of anomaly detection.

[0068] In some embodiments, the processor can also update the target knowledge graph based on real-time data; based on the updated target knowledge graph and preset trigger rules, determine whether to adjust the operation rules and whether to adjust the optimal operation plan.

[0069] The preset triggering rules may include season changes, weather changes, etc. As an example only, the preset triggering rule may be to automatically increase the frequency of landscape road cleaning during the leaf-falling season.

[0070] In some embodiments, the processor can adjust the operation rules and the optimal operation plan when the initial trigger rules are met. For example, during the leaf-falling season, the cleaning frequency of Nanjing Wutong Avenue is changed from once per day to three times per day.

[0071] In some embodiments, in response to the need to adjust the optimal job plan, the processor can determine the updated job scheduling sequence based on the processing of the updated target knowledge graph by the large language model; and determine the updated optimal job plan based on the processing of the updated job scheduling sequence by the optimization algorithm.

[0072] How to determine the updated job scheduling sequence and the updated optimal job plan is similar to the above steps 220 and 230. Figure 2 corresponding instructions.

[0073] As an example only, after receiving real-time data such as a rainstorm warning, the processor can output an updated emergency plan such as starting work two hours earlier and dispatching a vacuum truck to replace a sweeper.

[0074] The technical solution of the present invention realizes the upgrade of sanitation operations from "experience-driven" to "AI-driven", significantly improving scheduling efficiency and anomaly detection accuracy.

[0075] While the basic concepts have been described above, it will be apparent to those skilled in the art that the detailed disclosure is merely illustrative and does not limit this specification. Although not explicitly stated herein, various modifications, improvements, and revisions to this specification may be made by those skilled in the art. Such modifications, improvements, and revisions are suggested in this specification and remain within the spirit and scope of the exemplary embodiments of this specification.

[0076] This specification also uses specific terms to describe the embodiments of this specification. For example, "one embodiment," "an embodiment," and / or "some embodiments" refer to a feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "one embodiment," "an embodiment," or "an alternative embodiment" two or more times in different locations in this specification do not necessarily refer to the same embodiment. Furthermore, certain features, structures, or characteristics of one or more embodiments of this specification may be appropriately combined.

[0077] Similarly, it should be noted that, in order to simplify the presentation of this specification and thus facilitate understanding of one or more embodiments of the invention, the foregoing descriptions of the embodiments of this specification sometimes combine multiple features into a single embodiment, figure, or description thereof. However, this disclosure method does not imply that the subject matter of this specification requires more features than those recited in the claims. In fact, an embodiment may have fewer features than all of the features of a single disclosed embodiment.

[0078] Finally, it should be understood that the embodiments described in this specification are intended only to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, by way of example and not limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly described and illustrated in this specification.

Claims

1. A sanitation operation management and control method, characterized in that: include: Build a target knowledge graph based on historical data and / or environmental data; Determine a job scheduling sequence based on processing the target knowledge graph by a large language model; An optimal operation plan is determined based on processing of the operation scheduling sequence by an optimization algorithm.

2. The method according to claim 1, characterized in that The target knowledge graph includes one or more of a static graph, a dynamic graph, and an experience graph.

3. The method according to claim 1, characterized in that The processing of the job scheduling sequence based on the optimization algorithm to determine the optimal job solution includes: Determine the event urgency, resource consumption, and abnormality probability corresponding to the job scheduling sequence; Based on the event urgency, resource consumption, and abnormality probability, determining the reward value corresponding to the job scheduling sequence through the optimization algorithm; The optimal operation plan is determined based on the reward value.

4. The method according to claim 1, wherein The method further comprises: Obtain vibration data of work vehicles; Determining a false operation probability based on processing of the vibration data by an LSTM network; Obtaining images before and after the operation, and determining the similarity between the images before and after the operation; Based on the similarity between the images before and after the operation and the false operation probability, it is determined whether to trigger an early warning.

5. The method according to claim 4, characterized in that The method further comprises: Obtaining the operating time of the operating vehicle; Whether to trigger an early warning is determined based on the operation time, the similarity between the images before and after the operation, and the false operation probability.

6. The method according to claim 1, characterized in that The method further comprises: Updating the target knowledge graph based on real-time data; Based on the updated target knowledge graph and preset trigger rules, determine whether to adjust the operation rules and whether to adjust the optimal operation plan.

7. The method according to claim 6, characterized in that The method further comprises: Adjust the optimal operating plan as needed: Determining an updated job scheduling sequence based on processing the updated target knowledge graph by a large language model; Based on the processing of the updated job scheduling sequence by the optimization algorithm, the updated optimal job plan is determined.

8. A sanitation operation management and control system, characterized in that: include A graph construction module is configured to: construct a target knowledge graph based on historical data and / or environmental data; A model processing module is configured to: determine a job scheduling sequence based on processing of the target knowledge graph by a large language model; The solution determination module is configured to determine the optimal operation solution based on processing the operation scheduling sequence by an optimization algorithm.

9. A sanitation operation control device, characterized in that: The method comprises a processor configured to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The storage medium stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes the method according to any one of claims 1 to 7.

Images